Search

Recent works have been exploring the scaling laws in the field of Embodied AI. Given the prohibitive costs of collecting real-world data, we believe the Simulation-to-Real (Sim2Real) paradigm is a crucial step for scaling the learning of embodied models. This paper introduces project GRUtopia, the first simulated interactive 3D society designed for various robots. It features several advancements: (a) The scene dataset, GRScenes, includes 100k interactive, finely annotated scenes, which can be freely combined into city-scale environments. In contrast to previous works mainly focusing on home, GRScenes covers 89 diverse scene categories, bridging the gap of service-oriented environments where general robots would be initially deployed. (b) GRResidents, a Large Language Model (LLM) driven Non-Player Character (NPC) system that is responsible for social interaction, task generation, and task assignment, thus simulating social scenarios for embodied AI applications. (c) The benchmark, GRBench, supports various robots but focuses on legged robots as primary agents and poses moderately challenging tasks involving Object Loco-Navigation, Social Loco-Navigation, and Loco-Manipulation. We hope that this work can alleviate the scarcity of high-quality data in this field and provide a more comprehensive assessment of Embodied AI research. The project is available at https://github.com/OpenRobotLab/GRUtopia.

Object-oriented embodied navigation aims to locate specific objects, defined by category or depicted in images. Existing methods often struggle to generalize to open vocabulary goals without extensive training data. While recent advances in Vision-Language Models (VLMs) offer a promising solution by extending object recognition beyond predefined categories, efficient goal-oriented exploration becomes more challenging in an open vocabulary setting. We introduce OVExp, a learning-based framework that integrates VLMs for Open-Vocabulary Exploration. OVExp constructs scene representations by encoding observations with VLMs and projecting them onto top-down maps for goal-conditioned exploration. Goals are encoded in the same VLM feature space, and a lightweight transformer-based decoder predicts target locations while maintaining versatile representation abilities. To address the impracticality of fusing dense pixel embeddings with full 3D scene reconstruction for training, we propose constructing maps using low-cost semantic categories and transforming them into CLIP's embedding space via the text encoder. The simple but effective design of OVExp significantly reduces computational costs and demonstrates strong generalization abilities to various navigation settings. Experiments on established benchmarks show OVExp outperforms previous zero-shot methods, can generalize to diverse scenes, and handle different goal modalities.

Efficient finetuning of large language models (LLMs) aims to adapt the LLMs with reduced computation and memory cost. Previous LoRA-based approaches initialize the low-rank matrices with gaussian distribution and zero values, while keeping the original weight matrices frozen. However, the trainable model parameters optimized in an unguided subspace might have interference with the well-learned subspace of the pretrained weight matrix. In this paper, we propose MiLoRA, a simple yet effective LLM finetuning approach that only updates the minor singular components of the weight matrix while keeping the principle singular components frozen. It is observed that the minor matrix corresponds to the noisy or long-tail information, while the principle matrix contains important knowledge. The MiLoRA initializes the low-rank matrices within a subspace that is orthogonal to the principle matrix, thus the pretrained knowledge is expected to be well preserved. During finetuning, MiLoRA makes the most use of the less-optimized subspace for learning the finetuning dataset. Extensive experiments on commonsense reasoning, math reasoning and instruction following benchmarks present the superior performance of our method.

Fine-tuning is a crucial process for adapting large language models (LLMs) to diverse applications. In certain scenarios, such as multi-tenant serving, deploying multiple LLMs becomes necessary to meet complex demands. Recent studies suggest decomposing a fine-tuned LLM into a base model and corresponding delta weights, which are then compressed using low-rank or low-bit approaches to reduce costs. In this work, we observe that existing low-rank and low-bit compression methods can significantly harm the model performance for task-specific fine-tuned LLMs (e.g., WizardMath for math problems). Motivated by the long-tail distribution of singular values in the delta weights, we propose a delta quantization approach using mixed-precision. This method employs higher-bit representation for singular vectors corresponding to larger singular values. We evaluate our approach on various fine-tuned LLMs, including math LLMs, code LLMs, chat LLMs, and even VLMs. Experimental results demonstrate that our approach performs comparably to full fine-tuned LLMs, surpassing both low-rank and low-bit baselines by a considerable margin. Additionally, we show that our method is compatible with various backbone LLMs, such as Llama-2, Llama-3, and Mistral, highlighting its generalizability., Comment: 12 pages

LoRA employs lightweight modules to customize large language models (LLMs) for each downstream task or domain, where different learned additional modules represent diverse skills. Combining existing LoRAs to address new tasks can enhance the reusability of learned LoRAs, particularly beneficial for tasks with limited annotated data. Most prior works on LoRA combination primarily rely on task-level weights for each involved LoRA, making different examples and tokens share the same LoRA weights. However, in generative tasks, different tokens may necessitate diverse skills to manage. Taking the Chinese math task as an example, understanding the problem description may depend more on the Chinese LoRA, while the calculation part may rely more on the math LoRA. To this end, we propose LoRA-Flow, which utilizes dynamic weights to adjust the impact of different LoRAs. The weights at each step are determined by a fusion gate with extremely few parameters, which can be learned with only 200 training examples. Experiments across six generative tasks demonstrate that our method consistently outperforms baselines with task-level fusion weights. This underscores the necessity of introducing dynamic fusion weights for LoRA combination., Comment: Work in Progress

Stylistic headline generation is the task to generate a headline that not only summarizes the content of an article, but also reflects a desired style that attracts users. As style-specific article-headline pairs are scarce, previous researches focus on unsupervised approaches with a standard headline generation dataset and mono-style corpora. In this work, we follow this line and propose StyleBART, an unsupervised approach for stylistic headline generation. Our method decorates the pretrained BART model with adapters that are responsible for different styles and allows the generation of headlines with diverse styles by simply switching the adapters. Different from previous works, StyleBART separates the task of style learning and headline generation, making it possible to freely combine the base model and the style adapters during inference. We further propose an inverse paraphrasing task to enhance the style adapters. Extensive automatic and human evaluations show that StyleBART achieves new state-of-the-art performance in the unsupervised stylistic headline generation task, producing high-quality headlines with the desired style., Comment: Findings of EMNLP 2023

VLN-CE is a recently released embodied task, where AI agents need to navigate a freely traversable environment to reach a distant target location, given language instructions. It poses great challenges due to the huge space of possible strategies. Driven by the belief that the ability to anticipate the consequences of future actions is crucial for the emergence of intelligent and interpretable planning behavior, we propose DREAMWALKER -- a world model based VLN-CE agent. The world model is built to summarize the visual, topological, and dynamic properties of the complicated continuous environment into a discrete, structured, and compact representation. DREAMWALKER can simulate and evaluate possible plans entirely in such internal abstract world, before executing costly actions. As opposed to existing model-free VLN-CE agents simply making greedy decisions in the real world, which easily results in shortsighted behaviors, DREAMWALKER is able to make strategic planning through large amounts of ``mental experiments.'' Moreover, the imagined future scenarios reflect our agent's intention, making its decision-making process more transparent. Extensive experiments and ablation studies on VLN-CE dataset confirm the effectiveness of the proposed approach and outline fruitful directions for future work., Comment: Accepted at ICCV 2023; Project page: https://github.com/hanqingwangai/Dreamwalker

Vision-language navigation is a task that requires an agent to follow instructions to navigate in environments. It becomes increasingly crucial in the field of embodied AI, with potential applications in autonomous navigation, search and rescue, and human-robot interaction. In this paper, we propose to address a more practical yet challenging counterpart setting - vision-language navigation in continuous environments (VLN-CE). To develop a robust VLN-CE agent, we propose a new navigation framework, ETPNav, which focuses on two critical skills: 1) the capability to abstract environments and generate long-range navigation plans, and 2) the ability of obstacle-avoiding control in continuous environments. ETPNav performs online topological mapping of environments by self-organizing predicted waypoints along a traversed path, without prior environmental experience. It privileges the agent to break down the navigation procedure into high-level planning and low-level control. Concurrently, ETPNav utilizes a transformer-based cross-modal planner to generate navigation plans based on topological maps and instructions. The plan is then performed through an obstacle-avoiding controller that leverages a trial-and-error heuristic to prevent navigation from getting stuck in obstacles. Experimental results demonstrate the effectiveness of the proposed method. ETPNav yields more than 10% and 20% improvements over prior state-of-the-art on R2R-CE and RxR-CE datasets, respectively. Our code is available at https://github.com/MarSaKi/ETPNav., Comment: Project page: https://github.com/MarSaKi/ETPNav

Multilingual pretrained language models (mPLMs) have shown their effectiveness in multilingual word alignment induction. However, these methods usually start from mBERT or XLM-R. In this paper, we investigate whether multilingual sentence Transformer LaBSE is a strong multilingual word aligner. This idea is non-trivial as LaBSE is trained to learn language-agnostic sentence-level embeddings, while the alignment extraction task requires the more fine-grained word-level embeddings to be language-agnostic. We demonstrate that the vanilla LaBSE outperforms other mPLMs currently used in the alignment task, and then propose to finetune LaBSE on parallel corpus for further improvement. Experiment results on seven language pairs show that our best aligner outperforms previous state-of-the-art models of all varieties. In addition, our aligner supports different language pairs in a single model, and even achieves new state-of-the-art on zero-shot language pairs that does not appear in the finetuning process., Comment: Published at Findings of EMNLP 2022

With the emergence of varied visual navigation tasks (e.g, image-/object-/audio-goal and vision-language navigation) that specify the target in different ways, the community has made appealing advances in training specialized agents capable of handling individual navigation tasks well. Given plenty of embodied navigation tasks and task-specific solutions, we address a more fundamental question: can we learn a single powerful agent that masters not one but multiple navigation tasks concurrently? First, we propose VXN, a large-scale 3D dataset that instantiates four classic navigation tasks in standardized, continuous, and audiovisual-rich environments. Second, we propose Vienna, a versatile embodied navigation agent that simultaneously learns to perform the four navigation tasks with one model. Building upon a full-attentive architecture, Vienna formulates various navigation tasks as a unified, parse-and-query procedure: the target description, augmented with four task embeddings, is comprehensively interpreted into a set of diversified goal vectors, which are refined as the navigation progresses, and used as queries to retrieve supportive context from episodic history for decision making. This enables the reuse of knowledge across navigation tasks with varying input domains/modalities. We empirically demonstrate that, compared with learning each visual navigation task individually, our multitask agent achieves comparable or even better performance with reduced complexity., Comment: Accepted to NeurIPS 2022; Code: https://github.com/hanqingwangai/VXN

Since the rise of vision-language navigation (VLN), great progress has been made in instruction following -- building a follower to navigate environments under the guidance of instructions. However, far less attention has been paid to the inverse task: instruction generation -- learning a speaker~to generate grounded descriptions for navigation routes. Existing VLN methods train a speaker independently and often treat it as a data augmentation tool to strengthen the follower while ignoring rich cross-task relations. Here we describe an approach that learns the two tasks simultaneously and exploits their intrinsic correlations to boost the training of each: the follower judges whether the speaker-created instruction explains the original navigation route correctly, and vice versa. Without the need of aligned instruction-path pairs, such cycle-consistent learning scheme is complementary to task-specific training targets defined on labeled data, and can also be applied over unlabeled paths (sampled without paired instructions). Another agent, called~creator is added to generate counterfactual environments. It greatly changes current scenes yet leaves novel items -- which are vital for the execution of original instructions -- unchanged. Thus more informative training scenes are synthesized and the three agents compose a powerful VLN learning system. Extensive experiments on a standard benchmark show that our approach improves the performance of various follower models and produces accurate navigation instructions., Comment: Accepted to CVPR 2022

In cell-free massive MIMO networks, an efficient distributed detection algorithm is of significant importance. In this paper, we propose a distributed expectation propagation (EP) detector for cell-free massive MIMO. The detector is composed of two modules, a nonlinear module at the central processing unit (CPU) and a linear module at the access point (AP). The turbo principle in iterative decoding is utilized to compute and pass the extrinsic information between modules. An analytical framework is then provided to characterize the asymptotic performance of the proposed EP detector with a large number of antennas. Simulation results will show that the proposed method outperforms the distributed detectors in terms of bit-error-rate., Comment: 6 pages, 5 figures, 2 tables, Accepted by IEEE Globecom 2021

Scene Rearrangement Planning (SRP) is an interior task proposed recently. The previous work defines the action space of this task with handcrafted coarse-grained actions that are inflexible to be used for transforming scene arrangement and intractable to be deployed in practice. Additionally, this new task lacks realistic indoor scene rearrangement data to feed popular data-hungry learning approaches and meet the needs of quantitative evaluation. To address these problems, we propose a fine-grained action definition for SRP and introduce a large-scale scene rearrangement dataset. We also propose a novel learning paradigm to efficiently train an agent through self-playing, without any prior knowledge. The agent trained via our paradigm achieves superior performance on the introduced dataset compared to the baseline agents. We provide a detailed analysis of the design of our approach in our experiments., Comment: 7 pages, 4 figures

Recently, numerous algorithms have been developed to tackle the problem of vision-language navigation (VLN), i.e., entailing an agent to navigate 3D environments through following linguistic instructions. However, current VLN agents simply store their past experiences/observations as latent states in recurrent networks, failing to capture environment layouts and make long-term planning. To address these limitations, we propose a crucial architecture, called Structured Scene Memory (SSM). It is compartmentalized enough to accurately memorize the percepts during navigation. It also serves as a structured scene representation, which captures and disentangles visual and geometric cues in the environment. SSM has a collect-read controller that adaptively collects information for supporting current decision making and mimics iterative algorithms for long-range reasoning. As SSM provides a complete action space, i.e., all the navigable places on the map, a frontier-exploration based navigation decision making strategy is introduced to enable efficient and global planning. Experiment results on two VLN datasets (i.e., R2R and R4R) show that our method achieves state-of-the-art performance on several metrics., Comment: Accepted on CVPR2021; Implementation will be available at https://github.com/HanqingWangAI/SSM-VLN

Vision-language navigation (VLN) is the task of entailing an agent to carry out navigational instructions inside photo-realistic environments. One of the key challenges in VLN is how to conduct a robust navigation by mitigating the uncertainty caused by ambiguous instructions and insufficient observation of the environment. Agents trained by current approaches typically suffer from this and would consequently struggle to avoid random and inefficient actions at every step. In contrast, when humans face such a challenge, they can still maintain robust navigation by actively exploring the surroundings to gather more information and thus make more confident navigation decisions. This work draws inspiration from human navigation behavior and endows an agent with an active information gathering ability for a more intelligent vision-language navigation policy. To achieve this, we propose an end-to-end framework for learning an exploration policy that decides i) when and where to explore, ii) what information is worth gathering during exploration, and iii) how to adjust the navigation decision after the exploration. The experimental results show promising exploration strategies emerged from training, which leads to significant boost in navigation performance. On the R2R challenge leaderboard, our agent gets promising results all three VLN settings, i.e., single run, pre-exploration, and beam search., Comment: ECCV2020 (changed with improved perfromance on Pre-Explore and Beam Search settings); website: https://github.com/HanqingWangAI/Active_VLN

The combination of orthogonal frequency modulation (OFDM) and multiple-input multiple-output (MIMO) systems plays an important role in modern communication systems. In order to meet the growing throughput demands, future MIMO-OFDM receivers are expected to utilize a massive number of antennas, operate in dynamic environments, and explore high frequency bands, while satisfying strict constraints in terms of cost, power, and size. An emerging technology to realize massive MIMO receivers of reduced cost and power consumption is based on dynamic metasurface antennas (DMAs), which inherently implement controllable compression in acquisition. In this work we study the application of DMAs for MIMO-OFDM receivers operating with bit-constrained analog-to-digital converters (ADCs). We present a model for DMAs which accounts for the configurable frequency selective profile of its metamaterial elements, resulting in a spectrally flexible hybrid structure. We then exploit previous results in task-based quantization to show how DMAs can be configured to improve recovery in the presence of constrained ADCs, and propose methods for adjusting the DMA parameters based on channel state information. Our numerical results demonstrate that the DMA-based receiver is capable of accurately recovering OFDM signals. In particular, we show that by properly exploiting the spectral diversity of DMAs, notable performance gains are obtained over existing designs of conventional hybrid architectures, demonstrating the potential of DMAs for MIMO-OFDM setups in realizing high performance massive antenna arrays of reduced cost and power consumption.

The order-of-magnitude increase in the dimension of antenna arrays, which forms extra-large-scale massive multiple-input-multiple-output (MIMO) systems, enables substantial improvement in spectral efficiency, energy efficiency, and spatial resolution. However, practical challenges, such as excessive computational complexity and excess of baseband data to be transferred and processed, prohibit the use of centralized processing. A promising solution is to distribute baseband data from disjoint subsets of antennas into parallel processing procedures coordinated by a central processing unit. This solution is called subarray-based architecture. In this work, we extend the application of expectation propagation (EP) principle, which effectively balances performance and practical feasibility in conventional centralized MIMO detector design, to fit the subarray-based architecture. Analytical results confirm the convergence of the proposed iterative procedure and that the proposed detector asymptotically approximates Bayesian optimal performance under certain conditions. The proposed subarray-based EP detector is reduced to centralized EP detector when only one subarray exists. In addition, we propose additional strategies for further reducing the complexity and overhead of the information exchange between parallel subarrays and the central processing unit to facilitate the practical implementation of the proposed detector. Simulation results demonstrate that the proposed detector achieves numerical stability within few iterations and outperforms its counterparts., Comment: 15 pages, 16 figures, accepted by IEEE Transactions on Wireless Communications for publication

3D object reconstruction is a fundamental task of many robotics and AI problems. With the aid of deep convolutional neural networks (CNNs), 3D object reconstruction has witnessed a significant progress in recent years. However, possibly due to the prohibitively high dimension of the 3D object space, the results from deep CNNs are often prone to missing some shape details. In this paper, we present an approach which aims to preserve more shape details and improve the reconstruction quality. The key idea of our method is to leverage object mask and pose estimation from CNNs to assist the 3D shape learning by constructing a probabilistic single-view visual hull inside of the network. Our method works by first predicting a coarse shape as well as the object pose and silhouette using CNNs, followed by a novel 3D refinement CNN which refines the coarse shapes using the constructed probabilistic visual hulls. Experiment on both synthetic data and real images show that embedding a single-view visual hull for shape refinement can significantly improve the reconstruction quality by recovering more shapes details and improving shape consistency with the input image., Comment: 11 pages

The use of low-resolution analog-to-digital converters (ADCs) can significantly reduce power consumption and hardware cost. However, their resulting severe nonlinear distortion makes reliable data transmission challenging. For orthogonal frequency division multiplexing (OFDM) transmission, the orthogonality among subcarriers is destroyed. This invalidates conventional OFDM receivers relying heavily on this orthogonality. In this study, we move on to quantized OFDM (Q-OFDM) prototyping implementation based on our previous achievement in optimal Q-OFDM detection. First, we propose a novel Q-OFDM channel estimator by extending the generalized Turbo (GTurbo) framework formerly applied for optimal detection. Specifically, we integrate a type of robust linear OFDM channel estimator into the original GTurbo framework and derive its corresponding extrinsic information to guarantee its convergence. We also propose feasible schemes for automatic gain control, noise power estimation, and synchronization. Combined with the proposed inference algorithms, we develop an efficient Q-OFDM receiver architecture. Furthermore, we construct a proof-of-concept prototyping system and conduct over-the-air (OTA) experiments to examine its feasibility and reliability. This is the first work that focuses on both algorithm design and system implementation in the field of low-resolution quantization communication. The results of the numerical simulation and OTA experiment demonstrate that reliable data transmission can be achieved., Comment: 14 pages, 17 figures; accepted by IEEE Transactions on Communications

Machine learning (ML) has been widely applied to the upper layers of wireless communication systems for various purposes, such as deployment of cognitive radio and communication network. However, its application to the physical layer is hampered by sophisticated channel environments and limited learning ability of conventional ML algorithms. Deep learning (DL) has been recently applied for many fields, such as computer vision and natural language processing, given its expressive capacity and convenient optimization capability. The potential application of DL to the physical layer has also been increasingly recognized because of the new features for future communications, such as complex scenarios with unknown channel models, high speed and accurate processing requirements; these features challenge conventional communication theories. This paper presents a comprehensive overview of the emerging studies on DL-based physical layer processing, including leveraging DL to redesign a module of the conventional communication system (for modulation recognition, channel decoding, and detection) and replace the communication system with a radically new architecture based on an autoencoder. These DL-based methods show promising performance improvements but have certain limitations, such as lack of solid analytical tools and use of architectures that are specifically designed for communication and implementation research, thereby motivating future research in this field., Comment: Accepted by China Communications 14 pages, 19 figures

Orthogonal frequency division multiplexing (OFDM) has been widely used in communication systems operating in the millimeter wave (mmWave) band to combat frequency-selective fading and achieve multi-Gbps transmissions, such as IEEE 802.15.3c and IEEE 802.11ad. For mmWave systems with ultra high sampling rate requirements, the use of low-resolution analog-to-digital converters (ADCs) (i.e., 1-3 bits) ensures an acceptable level of power consumption and system costs. However, orthogonality among sub-channels in the OFDM system cannot be maintained because of the severe non-linearity caused by low-resolution ADC, which renders the design of data detector challenging. In this study, we develop an efficient algorithm for optimal data detection in the mmWave OFDM system with low-resolution ADCs. The analytical performance of the proposed detector is derived and verified to achieve the fundamental limit of the Bayesian optimal design. On the basis of the derived analytical expression, we further propose a power allocation (PA) scheme that seeks to minimize the average symbol error rate. In addition to the optimal data detector, we also develop a feasible channel estimation method, which can provide high-quality channel state information without significant pilot overhead. Simulation results confirm the accuracy of our analysis and illustrate that the performance of the proposed detector in conjunction with the proposed PA scheme is close to the optimal performance of the OFDM system with infinite-resolution ADC., Comment: 32 pages, 12 figures; accepted by IEEE JSAC special issue on millimeter wave communications for future mobile networks